DunnBC22's picture
Update README.md
1b02e1d
metadata
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
  - recall
  - precision
model-index:
  - name: ernie-2.0-base-en-Tweet_About_Disaster_Or_Not
    results: []
language:
  - en
license: cc

ernie-2.0-base-en-Tweet_About_Disaster_Or_Not

This model is a fine-tuned version of nghuyong/ernie-2.0-base-en on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2292
  • Accuracy: 0.9156
  • F1: 0.7876
  • Recall: 0.8436
  • Precision: 0.7386

Model description

This is a binary classification model to determine if tweet input samples are about a disaster or not.

For more information on how it was created, check out the following link: https://github.com/DunnBC22/NLP_Projects/blob/main/Binary%20Classification/Transformer%20Comparison/Is%20This%20Tweet%20Referring%20to%20a%20Disaster%20or%20Not%3F%20-%20ERNIE.ipynb

Associated Projects

This project is part of a comparison of multiple transformers. The others can be found at the following links:

Intended uses & limitations

This model is intended to demonstrate my ability to solve a complex problem using technology.

The main limitation is the quality of the data source.

Training and evaluation data

Dataset Source: https://www.kaggle.com/datasets/vstepanenko/disaster-tweets

Input Word Length By Class:

Length of Input Text (in Words) By Class

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 2e-05
  • train_batch_size: 64
  • eval_batch_size: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 5

Training results

Training Loss Epoch Step Validation Loss Accuracy F1 Recall Precision
0.347 1.0 143 0.2663 0.8777 0.7342 0.9100 0.6154
0.2192 2.0 286 0.2292 0.9156 0.7876 0.8436 0.7386
0.132 3.0 429 0.2629 0.9129 0.7843 0.8531 0.7258
0.0767 4.0 572 0.3266 0.9120 0.7807 0.8436 0.7265
0.0532 5.0 715 0.3622 0.9120 0.7788 0.8341 0.7303

Framework versions

  • Transformers 4.26.1
  • Pytorch 1.13.1
  • Datasets 2.9.0
  • Tokenizers 0.12.1